ESTRO 2024 - Abstract Book

S4362

Physics - Intra-fraction motion management and real-time adaptive radiotherapy

ESTRO 2024

The differences in volume (in absolute value) of prostate and VVSS in each adaptation of the plan were analyzed with respect to the reference volume contoured by the RO in the pCT. A U Mann-Whitney test was performed to compare differences among contouring methods for both structures. Statistical significance was set at p=0.05.

Results:

The differences in absolute value in the volume of the structures with the different methods are detailed. For the prostate, these values have been 29± 14% and 1,3±3,6 % for contouring by AI and RP, respectively. For the VVSS, they were 17±15%, 13±24% and 7±6% for the AI, RP and DP methods, respectively.

The differences obtained by comparing the different contouring methods were all statistically significant. For prostate, the p-value for the test of AI vs RP groups was p <0,001. For SSVV, the p-value was p = 0,039 for AI vs DP and p <0,001 for both AI vs DP and RP vs DP.

Thus, the variation in prostate contouring between the different fractions was appreciably reduced when using RP versus AI. This also occurred for VVSS, but was more pronounced when using PD without RP intervention.

Conclusion:

1. The reproducibility of prostate contouring is higher when using rigid prostate contouring propagation, so it can be considered to apply a margin reduction to PTV. 2. The reproducibility of prostate contouring is higher when using deformable propagation. A margin reduction can be considered when using this method.

In case we want to work with AI contours to reduce time during oART, the contour modification criteria should be clearly established to save time and maintain reproducibility between modifications of different ROs.

Keywords: Contours, reproducibility, on-line adaptive RT

2593

Digital Poster